AI Stocks' Wild Start to September: Mega-Rallies, Billion-Dollar Deals & Hype Under Fire

by Marcin Frąckiewicz
in Business, Finance, Technology
on 3 September 2025

Key Facts

Broad AI Sell-Off: After months of tech gains, the first trading days of September saw a notable pullback in AI-driven stocks. The Nasdaq and S&P 500 fell as investors rotated out of high-flyers. Nvidia dropped 3.5% on Tuesday – its steepest one-day decline in four months – amid a “broad reshuffling of risk” across crypto, high-beta tech, and AI plays reuters.com. Analysts say this looks less like panic and more like profit-taking as “valuations are high” and near-term catalysts have waned reuters.com.
Nvidia & Chipmakers Hit Headwinds: Nvidia’s stock slid to its lowest since late July, down about 8% since reporting Q2 results in August reuters.com. Rival chipmaker AMD traded lower in sympathy. Meanwhile, Washington tightened export rules: the U.S. revoked TSMC’s fast-track export license for chip equipment at its China fab (following similar moves for Samsung and SK Hynix) to curb China’s tech gains reuters.com reuters.com. The affected TSMC plant makes older 16nm chips (~2.4% of revenue), so impact is “minimal” – TSMC shares stayed flat even as SK Hynix and Samsung sank on the news reuters.com. U.S. chip-tool suppliers like KLA, Lam Research and Applied Materials face lower China sales due to the new licensing requirements reuters.com.
Microsoft’s $6B Government AI Deal: Microsoft struck a major agreement with the U.S. General Services Administration to deepen AI and cloud adoption across federal agencies. The deal – part of a White House push on AI – offers government-wide discounts on Azure cloud services and could save up to $3 billion in the first year reuters.com. Notably, Microsoft will grant free access to its new AI Copilot assistant for millions of existing government users for 12 months reuters.com, alongside reduced pricing on tools like Microsoft Sentinel security AI. (In August, GSA secured similar discounts from Google Cloud and AWS reuters.com.) Microsoft shares held steady on the news, as analysts see it cementing Microsoft’s role as a primary AI partner to Washington.
Alibaba’s $50B AI Rally in China: Chinese tech giant Alibaba saw its Hong Kong-listed stock surge 19% in one day, adding over $50 billion in value, after earnings revealed a “triple-digit” (100%+!) jump in AI-related product revenue and a 26% leap in cloud division sales businessmirror.com.ph businessmirror.com.ph. The blowout AI results helped offset concerns about Alibaba’s e-commerce price wars. “AI is delivering scalable growth, while traditional segments remain mired in price competition,” said Saxo strategist Charu Chanana, noting the “robust cloud sales show Alibaba is repositioning for longer-term relevance in the tech stack” businessmirror.com.ph. The news ignited China’s broader AI sector – Baidu jumped ~6% and Tencent also climbed in sympathy businessmirror.com.ph.
OpenAI’s Bold Moves (and $500B Ambition): ChatGPT-maker OpenAI announced it will acquire product-testing startup Statsig in an all-stock deal valuing Statsig at $1.1 billion, based on OpenAI’s own staggering ~$300 billion valuation reuters.com. As part of the deal, Statsig’s CEO Vijaye Raji becomes OpenAI’s new CTO of Applications, tasked with scaling ChatGPT and the Codex coding assistant amid intensifying competition reuters.com reuters.com. This comes just months after OpenAI’s $6.5 billion purchase of a Jony Ive-led AI hardware startup reuters.com. OpenAI’s revenues are soaring – it doubled annualized revenue to $12 billion in Jan–July and is on track for $20 billion by year-end reuters.com. The company is reportedly preparing a secondary share sale at a $500 billion valuation for investor liquidity reuters.com, underscoring how feverish demand for AI exposure has made OpenAI one of the world’s most richly valued startups.
Anthropic’s $13B Funding at $183B Value: In a testament to the global AI frenzy, rival AI lab Anthropic (maker of the Claude chatbot) raised a massive $13 billion Series F round that more than doubled its valuation to $183 billion reuters.com reuters.com. The round, led by investment firm ICONIQ Capital, marks a jump from Anthropic’s $62 billion valuation in March. Backers including Google and Amazon have fueled Anthropic’s explosive growth – the firm’s revenue run-rate rocketed from ~$1 billion at the start of 2025 to over $5 billion by August reuters.com. The cash infusion (co-led by Fidelity and Lightspeed, with participation from the Qatar Investment Authority, Blackstone, Coatue and others reuters.com) will fund scaling up infrastructure, “safety research” and global expansion reuters.com reuters.com. Analysts note that despite questions about high spending in AI, investor enthusiasm “stays strong” for top-tier startups reuters.com – U.S. venture funding in H1 2025 surged 75% YoY largely thanks to major AI bets reuters.com.
Palantir & the AI Valuation Debate: Data-analytics firm Palantir – now an “AI darling” of Wall Street – has seen its market cap skyrocket to about $372 billion, nearly matching the combined value of the three largest U.S. defense contractors reuters.com reuters.com. That price implies ~90× 2025 sales, making Palantir “one of the most expensive [stocks] in recent history,” according to Trivariate Research reuters.com. The stock’s meteoric rise (up ~140% YTD) has been fueled by AI buzz and meme-stock-like retail enthusiasm – retail investors plowed over $1.2 billion into Palantir shares in July alone reuters.com. Palantir’s business is real, however, and growing fast: revenue jumped 48% last quarter, and it recently won a $10 billion U.S. Army contract (over 10 years) and expanded a $1.3 billion AI defense program (Project Maven) reuters.com. Still, skeptics warn the stock is “priced to perfection” and question whether such hyper valuations can be sustained as the AI frenzy normalizes.

Market Overview: AI Stocks Face September Reality Check

The week opened with a notable cool-down in the red-hot AI trade, as investors returned from summer break and took profits in crowded tech positions. On September 2, U.S. indices fell roughly 0.7–0.8%, led by the previously high-flying “AI beneficiary” stocks reuters.com. This pullback followed a massive year-to-date rally in AI names and coincided with broader risk-off sentiment (amid worries over tariffs, inflation, and rising bond yields) reuters.com reuters.com.

Analysts and fund managers indicated that the selling was broad-based and largely driven by risk management rather than a fundamental crack in AI trends. “This week’s tech sell-off looks less like panic and more like a broad reshuffling of risk,” observed Bruno Schneller of Erlen Capital, noting that “crypto, high-beta tech and the AI beneficiaries all [came] under pressure at the same time” – a sign investors were cutting exposure across the board rather than reacting to any single AI news reuters.com. In other words, after months of “relentless upside,” the market hit a point where “we’ve run out of catalysts to buy more. Valuations are high. What can you point at to justify any higher?” as hedge fund manager Dan Izzo put it bluntly reuters.com.

Indeed, valuations for the AI leaders have grown rich after their huge run-ups. The Morgan Stanley AI Tech Index (a basket of top AI-exposed tech stocks) had jumped over 60% in 2025 through August, outpacing the broader market. But in the past two sessions it tumbled more than 5% – its worst two-day drop of the year – as some investors grew wary of “AI bubble” risk. Government bond yields near multi-year highs and seasonal factors (September is historically the weakest month for equities) also dampened risk appetite reuters.com. Citadel Securities strategists noted that stock buybacks and retail inflows, which had supported tech stocks over the summer, tend to “evaporate in September”, removing a key tailwind reuters.com.

Market Reaction: Prominent AI-heavy stocks all retreated to start the month. Nvidia (NVDA) – often viewed as the poster child of the AI boom – fell about 2% on Sept 1 and another 3.5% on Sept 2 reuters.com, touching its lowest levels since late July reuters.com. Even after this dip, Nvidia’s stock is up dramatically in 2025, but the post-earnings euphoria from August has given way to some profit-taking. Advanced Micro Devices (AMD), another key AI chip player, also slid roughly 2% over the two-day span (no major company-specific news, but it tends to trade in Nvidia’s slipstream). Among mega-cap tech, Microsoft, Alphabet (Google), Amazon, and Meta all saw modest declines (~1–2%) as part of the broader tech weakness. Electric vehicle leader Tesla, which has pitched itself as an AI company (with its self-driving tech), similarly lost over 1%.

Wall Street commentary suggests this “AI fatigue” may be a temporary reset rather than a trend reversal. Many analysts still see long-term upside in AI-oriented companies but caution that expectations had run too far, too fast. “Stocks of artificial intelligence companies really seem to be growing on a trend that’s going to be very difficult to sustain,” one strategist told Reuters reuters.com, arguing that near-term earnings will need to catch up to lofty share prices. The current pullback is giving the market a chance to digest recent gains and scrutinize actual AI business results going into year-end.

Chipmakers & AI Hardware: Nvidia Slips, Policy Risks Loom

High-performance semiconductor firms – which provide the “picks and shovels” of the AI revolution – were front and center in early-September trading and news. Nvidia, fresh off a historic rally that saw it briefly join the $1 trillion market cap club, experienced a rare stumble. By Sept 2, Nvidia’s stock had fallen ~8% from its post-earnings peak reuters.com. This came despite the company’s blowout Q2 results in late August, as some investors balked at the stock’s high valuation and as uncertainties emerged around its future growth rate.

One overhang: potential U.S.–China tech tensions that could restrict Nvidia’s access to a huge market for its AI chips. In a significant policy development, the Trump administration (in office as of 2025) moved to tighten export controls on chip technology to China – specifically by revoking “validated end-user” export licenses that previously allowed certain companies to ship U.S. chipmaking equipment to Chinese fabs without separate approvals reuters.com. On Sept 2, Washington rescinded Taiwan’s TSMC’s fast-track export status for its big Nanjing, China plant, just days after doing the same for South Korea’s Samsung and SK Hynix reuters.com.

This policy shift underscores U.S. efforts to prevent China from gaining cutting-edge semiconductor capabilities. However, the immediate impact on companies appears limited. TSMC’s Nanjing factory produces 16nm and other “mature node” chips – not TSMC’s most advanced nodes – and accounts for only ~2.4% of TSMC’s revenue reuters.com. TSMC said it is evaluating the situation but remains “committed to ensuring uninterrupted operations” in Nanjing reuters.com. Importantly, U.S. officials indicated they will grant licenses for foreign chipmakers to continue operating existing China facilities (maintenance, basic production), but not to expand or upgrade those plants’ tech levels reuters.com.

Market impact: Shares of Samsung Electronics and SK Hynix – which have large memory chip fabs in China – fell sharply on the license revocations. In contrast, TSMC’s stock barely budged (trading flat on Sept 3) since its China exposure is smaller and focused on older chip tech reuters.com. Semiconductor analysts noted the license changes could hit U.S. equipment suppliers like KLA, Lam Research, and Applied Materials, who may see reduced orders as Chinese fabs face new hurdles to importing tools reuters.com. On the flip side, Chinese domestic equipment-makers are not expected to gain much, given the complexity and the fact that most expansion plans in China were already near completion reuters.com.

Nvidia and AMD themselves are indirectly affected by these geopolitics. Both companies have had to sell curbed versions of their AI chips to China under earlier export rules, and there’s speculation the U.S. might tighten those rules further. Notably, in late August Nvidia’s CEO Jensen Huang downplayed fears of an AI slowdown, but did acknowledge that U.S. restrictions on selling to China create uncertainty in forecasts reuters.com reuters.com. Investors will be watching if Washington takes any new steps – for example, expanding the chips export ban – which could cut off what has been a major source of AI chip demand from Chinese cloud firms.

Meanwhile, the competitive landscape in AI hardware continues to evolve. AMD is positioning to challenge Nvidia’s dominance in AI accelerators: over the summer, CEO Lisa Su unveiled “MI300” series chips and a new AI supercomputing platform “Helios” aimed at large-scale data centers reuters.com reuters.com. AMD even brought OpenAI’s Sam Altman and Meta’s engineers on stage at its June event – a statement that major AI players are testing AMD’s alternatives reuters.com. The company has been on an acquisition spree (buying firms like Pensando, Nod.ai, and parts of Intel’s AI unit) to beef up its software and networking capabilities reuters.com reuters.com. While Nvidia remains far ahead, some analysts believe AMD could capture meaningful AI chip market share in 2024–26, especially if supply constraints or pricing pressure at Nvidia push cloud providers to seek options. Any traction by AMD could benefit its stock, which despite recent dips is still up ~40% YTD.

In summary, chip stocks are navigating a cross-current of sky-high AI demand on one hand and macro/policy risks on the other. Near term, profit-taking and export rules have introduced volatility. But longer term, the secular trend (massive investment in AI infrastructure globally) appears intact – a reason many investors are eager to buy these names on dips, assuming no severe escalation in U.S.–China trade frictions.

Cloud & Software Titans: Microsoft, Google, Amazon Double Down on AI

While chipmakers provide the raw horsepower for AI, the cloud and software giants are racing to be the platforms where that AI power is delivered to end-users and enterprises. Early this week, several major developments underscored how Big Tech is integrating AI into their strategies – and securing big customers in the process.

Microsoft (MSFT) grabbed headlines with a landmark deal to bring AI to the U.S. government. On Sept 2, Microsoft and the General Services Administration (GSA) announced a new agreement – dubbed “Project Accelerate” – aimed at infusing AI and cloud tools across federal agencies reuters.com. The deal implements a government-wide “unified pricing” approach that could save U.S. agencies up to $3 billion in the first year alone reuters.com. In practice, Microsoft is offering steep discounts on its Azure cloud services, cybersecurity tools, and notably free trials of Microsoft 365 Copilot – its flagship generative AI assistant – for millions of government workers for up to 12 months reuters.com. The White House has prioritized AI adoption (via its National AI Action Plan), and this partnership helps agencies experiment with AI-enhanced productivity software (like Copilot in Office apps) in a budget-friendly way.

Microsoft’s President Brad Smith called it a “win-win” that will accelerate AI use in a secure, compliant manner for public sector needs ranging from document summarization to threat analysis. For Microsoft, it virtually locks in the federal government as a long-term Azure/OpenAI customer. Analysts noted this mirrors the strategy in the private sector – hook users with AI features and then drive cloud consumption. The GSA had earlier negotiated similar cloud discount pacts with Google and Amazon Web Services (AWS) in August reuters.com, so Microsoft was actually playing catch-up to ensure Azure isn’t left out of federal AI workloads. Still, Microsoft’s inclusion of Copilot at no extra cost is a unique sweetener that could give it an edge as agencies evaluate AI solutions. The stock market viewed the news positively; MSFT shares were flat-to-slightly up, holding near $330 (~$2.5 trillion market cap). The deal reinforces Microsoft’s image as a trusted AI provider – an important signal as it competes with AWS and Google Cloud for enterprise and government clients.

Over at Google/Alphabet (GOOGL), the week brought a mix of legal relief and ongoing AI investments. On Sept 2, Alphabet’s stock jumped ~1.5% after a U.S. judge ruled in an antitrust case that Google can maintain certain search distribution deals – easing some regulatory concerns reuters.com. While not directly AI-related, this positive ruling boosts Google’s ability to fund its AI ambitions with healthy search profits. In the AI realm, Google is preparing to launch its next-gen model Gemini (seen as a GPT-4 competitor) and continues to integrate AI into flagship products like Search, Gmail, and Cloud. Google is also a key backer of Anthropic (now worth $183B after its raise reuters.com), which means Google’s cloud platform stands to benefit from Anthropic’s growth (Anthropic uses Google Cloud as a preferred provider and Google owns a reported ~10% stake). Furthermore, Google’s own cloud unit reportedly secured a GSA contract in August similar to Microsoft’s, offering AI tools under discounted terms, which positions Google AI offerings like Bard and its Vertex AI platform for federal use. While Google had no singular splashy announcement on Sept 2–3, its steady stream of AI updates – from rolling out new Duet AI features in Workspace to improving the Bard chatbot – keeps it firmly in the AI race. Investors are watching for Google’s next big AI event, and the stock’s resilience suggests confidence that Google will not be left behind in the “AI era.”

Amazon (AMZN), likewise, has been active on multiple AI fronts. Though no new deal was announced this week (after its August GSA cloud discount offer reuters.com), reports continue to swirl that Amazon is considering a new multibillion-dollar investment in Anthropic to deepen their partnership reuters.com. AWS already offers Anthropic’s Claude model as part of its Bedrock AI service, and a bigger stake could ensure Amazon Alexa and AWS have priority access to Anthropic’s latest innovations. Amazon’s strategy is two-pronged: build its own AI chips (the Inferentia and Trainium silicon, to lower the cost of AI cloud services) and offer a “bring-your-own-model” smorgasbord so enterprise customers can choose AI models from Amazon or third parties on AWS. In late August, Amazon unveiled “AWS HealthScribe,” an AI tool for clinical documentation, and is reportedly working on more industry-specific AI solutions. Amazon’s stock, like its peers, dipped ~1% in early September trading, reflecting general tech sentiment rather than AI-specific issues.

It’s worth noting that Meta Platforms (META) is making waves by open-sourcing some AI capabilities – an approach divergent from the cloud pay-per-use model. Meta’s new large language model Llama 2 (launched in July) is freely available to developers and has seen strong uptake. On Sept 3, Meta’s CTO scoffed at rivals’ “closed” AI systems and hinted Meta might open-source an AI model for image generation next. Meta is betting that getting its AI models widely adopted (even free) will drive longer-term benefits for its ecosystem and ads business. While Meta had no major AI news on these specific days, its stock has been buoyed by optimism that AI can meaningfully boost engagement and ad targeting across Facebook, Instagram, and its new Threads platform.

Bottom line: The Big Tech titans are deeply entwining AI into their offerings and securing strategic deals that could yield rich dividends. Microsoft’s public-sector win and Amazon/Google’s cloud maneuvers show AI is now a key battleground in enterprise/government sales. Each company is playing to its strengths – Microsoft leveraging its software suite and OpenAI alliance, Google leaning on its search/data prowess and research, Amazon on its cloud dominance and commerce data, and Meta on its social platforms and open-source approach. For investors, this means the “Magnificent Seven” tech stocks will likely continue to trade on every AI-related update. Expect more announcements (and perhaps new partnerships or acquisitions) in the coming weeks as these giants vie for AI leadership across cloud, consumer, and enterprise domains.

China’s AI Surge: Alibaba’s Blockbuster Earnings and Beyond

Halfway around the world, China’s tech sector offered a very different AI narrative: one of accelerating growth and investor excitement rather than tempering hype. The standout story was Alibaba Group’s explosive rally following its quarterly earnings report that put to rest any doubts about Chinese firms’ ability to monetize AI. On September 1 (Asia time), Alibaba revealed that its investments in artificial intelligence are already yielding substantial returns:

The company’s AI-driven product lines saw revenue soar by “triple digits” (%) year-on-year businessmirror.com.ph. This likely includes Alibaba’s cloud AI services, its large language model offerings (such as Tongyi Qianwen, which Alibaba opened for public use in August), and AI integrations in e-commerce and marketing products.
Alibaba’s cloud computing division – which is China’s largest cloud provider and integral to AI – grew sales 26% YoY, handily beating expectations businessmirror.com.ph. Management credited surging demand from Chinese businesses deploying AI applications on Alibaba Cloud.

These stellar AI results helped drive Alibaba’s overall revenue and profit above forecasts. They also alleviated concerns about intense competition in China’s e-commerce market (where rivals JD.com and Meituan have been cutting prices). Investors reacted with an emphatic buying spree: Alibaba’s Hong Kong-listed shares rocketed +19% in a single session, the stock’s biggest jump since 2022, adding over $50 billion in market cap businessmirror.com.ph. Trading volume hit a record high as both local and global investors rushed in. In the U.S., Alibaba’s ADR (BABA) leapt about 13% when New York trading resumed after Labor Day, reflecting renewed confidence in the company’s post-restructuring prospects.

Crucially, Alibaba’s surge didn’t happen in isolation – it ignited a broader rally in Chinese “AI economy” stocks. Search leader Baidu, which has its own Ernie AI model, climbed ~6% in Hong Kong as the strong read-through from Alibaba suggested that AI adoption is boosting revenues across the sector businessmirror.com.ph. Gaming and social giant Tencent (investor in many AI startups) also gained ~3%. Smaller AI players, from cloud providers to chip-design startups in Shenzhen, saw lifts in their stock prices following Alibaba’s news. This dynamic underlines how, in China, AI is becoming a key differentiator for tech companies’ growth trajectories.

Analysts were quick to frame Alibaba’s results as a watershed moment. “Alibaba’s earnings underscore a bifurcation within China tech: AI is delivering scalable growth, while traditional consumer-facing segments remain mired in destructive price competition,” said Charu Chanana of Saxo Markets businessmirror.com.ph. The implication is that those Chinese companies who have invested early and aggressively in AI (Alibaba, Baidu, possibly Huawei in chips, etc.) are pulling ahead, while those sticking to legacy businesses are struggling. Chanana added that “the triple-digit surge in AI revenue and robust cloud sales show Alibaba is repositioning for longer-term relevance in the tech stack, not just retail dominance.” businessmirror.com.ph In other words, Alibaba is evolving from an e-commerce company into a full-spectrum tech infrastructure and AI provider – a narrative investors love, as it mirrors the high-multiple story of U.S. firms like Amazon or Microsoft.

It wasn’t all roses in China tech land – the backdrop includes a cooling economy and other firms facing challenges. For instance, Alibaba’s e-commerce peers JD.com and Meituan recently saw profits slump amid their price war (JD’s profit plunged 50%, Meituan warned of losses) businessmirror.com.ph. But tellingly, Alibaba’s AI and cloud strength “glossed over” those concerns businessmirror.com.ph, convincing investors that Alibaba can outgrow the competition through tech leadership. A Morgan Stanley note went so far as to call Alibaba “China’s best AI enabler” with a multi-year advantage businessmirror.com.ph.

Beyond Alibaba, China’s government and industry are heavily backing AI as a strategic sector. Over the past week, state media reported that several Chinese local governments (e.g. Beijing, Shenzhen) will offer subsidies for AI cloud computing usage and incentivize companies building large models. There is also buzz that the People’s Bank of China might fast-track IPO approvals for AI-focused companies to stimulate tech investment. Such support, combined with blockbuster corporate results, has created a stark contrast to the West: while U.S. markets fret about an AI bubble, China’s AI sector is in a boom phase of real revenue growth. As Chanana noted, “parts of China tech are quietly reaccelerating – driven not by hype, but by real revenue growth in AI and cloud… It’s not a broad-based rotation yet, but the divergence is real.” businessmirror.com.ph

For global investors, one caveat is that many Chinese AI winners (like Alibaba, Baidu, Tencent) are listed in Hong Kong and subject to different regulatory and geopolitical risks (e.g. U.S.–China relations, potential sanctions). But with valuations of Chinese tech still relatively depressed compared to U.S. peers, some international funds are starting to see opportunity. Alibaba’s American depositary shares, for example, still trade at under 12× forward earnings – a fraction of Amazon’s multiple – suggesting room for upside if Alibaba’s AI-fueled growth continues.

In sum, China’s AI landscape in early September 2025 is marked by tangible breakthroughs: companies are monetizing AI at scale, and the stock market is rewarding those results. This momentum could prompt even more aggressive investment in AI by Chinese firms (e.g. Tencent ramping up AI R&D, Baidu pushing Ernie 4.0 model launch by year-end, Alibaba pouring billions more into its semiconductor unit T-Head). It’s a space to watch, as China aims to challenge U.S. dominance in AI – and if these earnings are any indication, they are making swift progress.

Private AI Titans: OpenAI & Anthropic Attract Massive Capital

It’s not just public markets riding the AI wave – private venture and equity financing in the AI sector is reaching historic highs, as evidenced by two eye-popping deals disclosed on Sept 2:

Anthropic’s $13 Billion Fundraise: Anthropic, a San Francisco-based AI startup known for its Claude chatbot, announced it has secured $13 billion in new funding in a Series F round reuters.com. This catapulted Anthropic’s post-money valuation to $183 billion – more than double its $61.5 billion valuation from just March reuters.com. To put that in perspective, Anthropic is now worth more than SAP or Siemens, despite being a 4-year-old company. The round was led by ICONIQ Capital (a fund linked to Mark Zuckerberg and other Silicon Valley elite), with major participation from Fidelity, Lightspeed, Blackstone, Coatue, and sovereign wealth funds like the Qatar Investment Authority reuters.com. Notably, Anthropic’s strategic backers Google and Amazon also have stakes (Google invested ~$400M earlier; Amazon is contemplating a fresh multibillion infusion reuters.com).

Investors are effectively betting that Anthropic will be one of a few dominant “AI model providers” in the West, alongside OpenAI and possibly Google itself. Anthropic has set itself apart partly by focusing on AI safety and reliability and by excelling in AI coding capabilities reuters.com. Its Claude model is considered one of the top GPT-4 rivals. According to Reuters, Anthropic’s revenue trajectory is astonishing – from a ~$1B annual run-rate in January to over $5 billion in run-rate by August 2025 reuters.com. This suggests companies are paying big bucks for Claude’s services (likely via cloud deals with Google) and that Anthropic is rapidly signing on enterprise clients. The new $13B war chest will help Anthropic train its next-generation models (which will require enormous computing power), deploy infrastructure globally, and perhaps crucially meet the “growing enterprise demand” it is seeing reuters.com. In a blog, Anthropic said the funds will also deepen its research into AI safety as it scales up – a point meant to assure regulators and the public.

The sheer size of this raise underlines how AI startups now command valuations typically reserved for the biggest tech companies. Anthropic at $183B is valued higher than OpenAI’s last implied valuation (~$80–90B) and even above Meta’s market cap as of mid-2024. Some observers call it “AI Gold Rush 2.0” – reminiscent of the dot-com era’s later stages, but with revenue to back it up. PitchBook data shows U.S. startup funding overall jumped 75% in H1 2025, “driven largely by major AI investments” reuters.com. Clearly, deep-pocketed investors don’t want to miss out on the next OpenAI, and they are willing to write enormous checks despite high burn rates and competition.

OpenAI’s Acquisition Spree & $500B Ambition: Not to be outdone, OpenAI (creator of ChatGPT) made news on the M&A front. The company announced the acquisition of Statsig, a 3-year-old startup that builds tools for A/B testing and feature rollout (essentially helping developers test new app features) reuters.com reuters.com. OpenAI will pay for Statsig in an all-stock deal valuing the startup at ~$1.1 billion reuters.com – and notably, that valuation is based on OpenAI’s own sky-high internal valuation of $300B. In exchange, OpenAI gains a product-focused team led by former Facebook veteran Vijaye Raji, who will become OpenAI’s “Chief Technology Officer of Applications.” This is a new high-level role overseeing how OpenAI’s models are turned into consumer and business applications (like ChatGPT, the Codex coding assistant, possibly the rumored “AI assistant” device collaboration with Jony Ive, etc.) reuters.com reuters.com. CEO Sam Altman appears to be bolstering OpenAI’s ranks with seasoned execs (earlier this year it hired Alexa AI chief Matsuo, and acquired Ive’s hardware design startup in a $6.5B deal reuters.com) to ensure it can productize AI and fend off competition from tech giants and well-funded rivals like Anthropic.

Perhaps even more striking was the revelation – via Reuters reporting – that OpenAI is in talks to sell existing shares at a $500 billion valuation in a tender offer reuters.com. This would allow employees and early backers to cash out some stake at an eye-popping half-trillion valuation, and bring in new investors without issuing new shares. If it goes through, OpenAI would become one of the world’s most valuable private companies, on par with the likes of Tesla and Meta in market cap (and trailing only a handful like Apple, Microsoft, Aramco). Altman has hinted in the past that OpenAI’s goal is to eventually be “the most capital-intensive startup in Silicon Valley” as it seeks Artificial General Intelligence. Indeed, OpenAI’s capital needs are enormous – training GPT-5 and beyond, running giant models for millions of users (the company’s inference costs for ChatGPT are reportedly over $700k per day), and now hardware projects – so raising cash at favorable valuations makes sense. OpenAI’s revenues are growing exponentially (on pace for $1+ billion this year, by some reports), but likely not yet enough to cover all its expenses, thus necessitating external funding.

These moves by OpenAI and Anthropic underscore a key point: the center of gravity in AI is partly shifting from the public markets to the private realm, where flexibility and massive capital infusions allow rapid scaling. For public-market investors, it creates a dilemma – you can’t directly invest in OpenAI or Anthropic unless they IPO (and none have immediate plans to). Instead, investors gain exposure through the big tech firms that have stakes or partnerships: Microsoft (owns 49% of OpenAI’s economic rights and uses OpenAI tech in Azure), Google (owns ~10% of Anthropic, plus its cloud hosts Anthropic), Amazon (in talks to increase its Anthropic stake; also using its models), and to a lesser extent Salesforce, Zoom, and others that have alliances with these startups.

Another angle is M&A: We’re seeing the first real consolidation moves (OpenAI buying startups like Statsig and the AI design firm from Ive). It raises the question of whether giants like Microsoft or Google might eventually try to acquire one of these AI unicorns outright. However, given the valuations (hundreds of billions), a full takeover would be difficult and likely face antitrust pushback. Instead, strategic minority investments seem to be the route for now.

In summary, the private AI labs are flush with cash and racing ahead. Anthropic’s $13B round and OpenAI’s aggressive expansion show no signs of a funding winter for top-tier AI players – rather, an arms race to build the most advanced models and capture market share. How well they execute will determine if these valuations are justified. For now, investors appear willing to suspend disbelief and pour money into anything AI-related – a confidence that is buoying the entire AI ecosystem but also prompting some to caution that discipline and tangible results must follow all the money.

Sky-High Valuations Under Scrutiny: Palantir & the “AI Hype” Question

One of the most striking stock stories of 2025 has been the resurrection of Palantir Technologies (PLTR) as a self-described “AI company.” The data analytics firm – traditionally known for government software contracts – pivoted hard into AI this year with its “Artificial Intelligence Platform” and a string of buzzworthy press releases about military AI usage. The result: Palantir’s stock has exploded over 150% in 2025, making it the top-performing S&P 500 component at times. By early September, Palantir’s market cap hovered around $372 billion, putting it in league with the largest defense contractors on the planet reuters.com reuters.com.

Such a valuation implies dizzying multiples. Palantir trades at roughly 90× its expected 2025 sales, compared to about 2× sales for established defense giants like Lockheed Martin or Raytheon reuters.com. Even mega-cap growth stocks rarely sustain 90× revenue; as a Reuters Breakingviews analysis noted, “among ~2,000 sizable U.S. companies that have ever traded above a 70× multiple [since 2000], Palantir is one of the most expensive in recent history” reuters.com. This has raised eyebrows on Wall Street about whether AI euphoria has run ahead of fundamentals in Palantir’s case.

Fueling Palantir’s surge is a potent mix of retail investor enthusiasm and genuine business momentum. On the hype side, Palantir has become a retail trader darling (often trending on Reddit’s WallStreetBets). In July alone, individual investors poured $1.2 billion into Palantir stock – a meme-stock caliber inflow on par with names like Tesla reuters.com. Social media is rife with bullish takes on Palantir as “the next NVIDIA” or the future “operating system for AI in the military.” This fervor, of course, can cut both ways – it introduces volatility and the risk of sharp pullbacks if sentiment sours. Indeed, Palantir’s shares saw a 15% “mini-slump” at one point in August, reminding that stocks priced for perfection have no margin for error reuters.com.

On the fundamental side, Palantir’s recent results do show improvement. Q2 revenue grew 13% (a re-acceleration from prior quarters, thanks in part to AI platform sales), and crucially, Palantir reported its first-ever profitable quarter on a GAAP basis. More impressively, Palantir is guiding to 20%+ revenue growth next year, and as mentioned, its government and commercial AI projects are scaling. The company has snagged some enormous contracts: in the past few months it won a $10 billion, 10-year deal with the U.S. Army to be the backbone of the Army’s new digital infrastructure reuters.com. It also expanded a contract for the Pentagon’s Project Maven (which uses AI to analyze drone and satellite imagery) to $1.3 billion reuters.com. CEO Alex Karp has been on a media blitz highlighting how Palantir’s AI can “turn the tide” by integrating all sorts of battlefield data – effectively positioning Palantir as the premier defense AI provider. The company boasts that it is now the “prime contractor” (lead) on nine major U.S. Army programs of record reuters.com, something unheard of for a software firm, as primes are usually hardware makers like Boeing or Raytheon.

All this has lent credibility to Palantir’s claim that it can eventually generate tens of billions in revenue and justify its valuation. Some analysts, like those at Morningstar, argue that Palantir could capture 2–3% of a $1.6 trillion global software/AI market by 2030 – which would imply ~$45–50B revenue. That still wouldn’t fully justify the current $370B market cap (which implies ~$48B revenue at 7.6× sales, slightly richer than Microsoft’s multiple) reuters.com, but it comes closer. Moreover, Karp has teased that Palantir’s commercial (non-government) business – which includes its Foundry platform used by enterprises – could grow “10× by 2030.” In Q2, U.S. commercial revenue did jump 93% (thanks to demand for AI deployments in industries like aviation, auto, healthcare) reuters.com. If Palantir can indeed sustain, say, 35–40% annual growth for several years (a big “if”), then today’s valuation may be vindicated in hindsight.

For now, however, many market veterans urge caution. Dean Kisratti, a strategist at UBS, told clients that while Palantir is a unique asset, “the stock is priced beyond perfection… any hiccup in the AI narrative or slowdown in deal ramp-up could trigger a sharp correction.” Short interest in PLTR has crept up, indicating some are betting on a fall. We also saw a director at Palantir sell $50 million worth of shares in late August – an insider sale that some interpreted as taking profits at high levels.

In broader context, Palantir exemplifies the froth and promise of the AI boom: It has become a $300B+ company on AI hopes nearly overnight, yet it also has real technology and contracts making AI a reality in critical domains. The stock’s journey from here will likely depend on two things: execution and sentiment. If Palantir executes flawlessly – delivering the 30-40% growth and margin expansion bulls expect – it could “grow into” its valuation over time (much like Amazon did in the 2010s). But if the AI project wins slow down or competitors emerge (say, Microsoft or Google targeting defense deals), sentiment could sour quickly on the high multiple.

One key barometer will be Palantir’s upcoming earnings and whether the company’s new AI products (like its AIP platform, which allows users to chat with their private data using LLMs) start contributing meaningfully to revenue. Early indications are positive – Palantir said more than 100 organizations have signed up for its AI platform in just a couple months. The U.S. government’s budget trajectory is another factor: If defense AI spending keeps rising (the 2024 National Defense Authorization Act has billions earmarked for AI), Palantir stands to benefit disproportionately.

In conclusion, Palantir stands at the intersection of AI optimism and valuation anxiety. It’s a case study in how the stock market is trying to price the “AI revolution”: with a mix of exuberance and trepidation. As we move forward, expect more dramatic swings in Palantir and its peers (e.g. C3.ai, another enterprise AI stock) as each new contract or earnings report tests the durability of the AI hype. For investors, the lesson of early September is clear – distinguishing hype from reality in AI will be paramount, and even in a gold rush, one must beware of heights.

Industry & Analyst Reactions

The whirlwind of AI stock news over the past 48 hours drew a range of reactions from market analysts, tech industry figures, and financial commentators:

“Peak AI?” – Or a Healthy Pause: Many strategists saw the U.S. tech sell-off as a natural breather. “After a summer of strong positioning and relentless upside, September historically brings a shift,” noted Scott Rubner of Citadel Securities reuters.com. He and others pointed out that systematic funds had loaded up on tech, and now some are trimming. Still, few doubt the longer-term AI story. Dan Ives of Wedbush (one of Wall Street’s tech bulls) wrote that any near-term weakness in AI stocks is “a golden opportunity, not the end of the rally,” arguing that enterprise IT spending on AI is just beginning a multi-year cycle. On the other hand, Morgan Stanley’s Mike Wilson – known for skepticism – warned that parts of the market “may be priced for perfection on AI, setting up potential disappointment if growth doesn’t materialize by early 2024.”
Big Bank Views on Big Tech: In light of Microsoft’s government deal, analysts lauded the company’s execution. “Microsoft is putting a moat around its AI offerings,” said Bernstein’s Mark Moerdler, who sees the GSA deal as “a template for large enterprise contracts” to come reuters.com reuters.com. Over at Google, Citi’s analysts highlighted that Google’s investments in startups like Anthropic “validate its AI ecosystem strategy”, even if they raise spending. One intriguing quote came from Morgan Stanley’s note on Alibaba: “Alibaba’s breakout reinforces a broader theme… parts of China tech are quietly reaccelerating — driven not by hype, but by real revenue growth in AI and cloud,” wrote MS strategist Gary Yu businessmirror.com.ph, basically telling global investors not to sleep on Chinese AI leaders.
AI Startup Frenzy: Venture capital insiders described Anthropic’s $13B raise as “staggering”. A partner at Sequoia commented to the Financial Times, “We’re in uncharted territory. These AI companies are raising more in one round than the entire tech sector did a decade ago.” Nonetheless, there’s fear of missing out: with only a handful of foundational model companies out there, “everyone wants a horse in the race,” as one VC told Reuters. Sam Altman’s bold $500B valuation target for OpenAI also drew chatter. Some praised Altman’s vision, noting OpenAI’s revenue ramp and dominance with ChatGPT. Others, like tech columnist Tim O’Reilly, cautioned “a $500B tag would imply a >25× multiple on future revenue – even the biggest tech companies rarely hit that. OpenAI will need to unlock whole new markets to justify it.”
Voices of Caution: Amid the exuberance, a few veteran investors urged sobriety. Famed value investor Jeremy Grantham mused that the “AI bubble” reminds him of the dot-com era, adding “I believe AI will change the world, but the market may have already priced in years of growth.” Similarly, Dean Turner of UBS said in a Reuters interview, “Stocks of artificial intelligence companies really seem to be growing on a trend that’s going to be very difficult to sustain”, pointing to Palantir as an example of rich valuations built on future hopes reuters.com. Notably, short-seller Jim Chanos revealed he is short some AI-exposed semiconductor names, arguing that AI demand, while strong, “won’t prevent cyclicality in the chip business.” It remains to be seen if these contrarian bets pay off; so far in 2025, doubting the AI boom has been a losing trade.
Tech CEOs on AI: Finally, we heard from some industry players. NVIDIA’s CEO Jensen Huang (fresh off selling some personal stock for hundreds of millions) insisted in a panel on Wednesday that “the AI boom is far from over – we’re just in the first inning”, dismissing concerns of a slowdown reuters.com. Elon Musk – who recently started his own AI venture xAI – tweeted that the rapid fundings of OpenAI and Anthropic show “AI competition is heating up; need to move faster at xAI.” Musk also made a provocative comment that “if AI companies can fetch $200B-$300B valuations pre-AGI, imagine when we get AGI”, hinting at even larger future stakes. Meanwhile, Meta’s Mark Zuckerberg said on an internal call (according to a leak) that Meta’s open-source AI approach “will win in the long run by driving innovation everywhere, even if Wall Street doesn’t get it yet.”

Overall, the sentiment can be summed up in a quote by Charu Chanana (Saxo Markets) after Alibaba’s earnings: “This isn’t a broad-based rotation yet – but the divergence is real.” businessmirror.com.ph In other words, the AI boom is creating winners and losers, and savvy investors/experts are trying to discern which moves are grounded in real growth (like Alibaba’s AI revenue) versus which are purely multiple expansion. The past two days of news gave fodder to both sides of the debate. Going forward, expect the tug-of-war between AI optimism and valuation concerns to continue defining market narratives.

Sources: